Comparing Arrays in same dataset?

[Reeza]

array1 should be the smaller list. 

This doesn't account for missing though so you may want to add a filter for missing to stop the DO loop.

[shellp55]

Got it, makes sense because it is matching blank to blank and considers that a match.  Sorry for being dense but not sure how to add that filter...but I will search the site.  Thanks again for all your help.

[Reeza]

if missing(_proc1(i)) then leave;

I think you add that into the top of the do loop. 

This does assume that once you find a missing value all other values will be missing, which is pretty much how I recall medical data being formatted.

[Reeza]

It's leave or break...you can test them 😉

[shellp55]

Thanks Reeza!  I've tried the code in various spots but it still shows as zero cases and I know there are some matching the criteria.

I've used:

do I = 1 to dim(_proc1);
find_index=whichc(_proc1(i) of _proc2(*));
if missing(_proc1(i)) then leave;
else if find_index>0 then do;
if _date1(i)=_date2(find_index) then flag=1;
end;
end;

[Reeza]

Change your sample data to better reflect your actual data and then I can test it. Otherwise, I'm guessing.

[Reeza]

@shellp55 You can also email me the sample data if you don't want to post it publically. I *think* I've worked with you before as a consultant and have a privacy contract that would still cover this. 

[shellp55]

Thanks Reeza but it will take awhile to de-identify my data.  I think you are correct that it is the blank values because I changed my test data so that there is no data in _proc1 and _date1 and it erroneously flagged the case as matching my criteria.  i'm trying to use if not missing(px1) to include an 'else if missing(_proc1(i)) then leave but I get an error that I do not have a matching do/select statement.